Thread by @fuserleer

Thanks #ftx, #dex vs #cex debate is hot again in #crypto

I've spent many years thinking about #dex implementations & building prototypes.

Lets dive in a little & explore the possibility of replacing #exchanges with #decentralized alternatives today

A small thread 🧵

1) Firstly we need throughput, lots & lots of it. A popular exchange such as @binance is likely performing 10,000s of trades per second across 100s of pairs.

A few years ago this kind of throughput was a dream, but not so much today 💪

Our consensus mechanism at #radix could handle that level of throughput, other projects are also approaching that tipping point too.

2) Finality is second. All #dex implementations so far have been low frequency trading only, partly because of throughput, but also because of the consensus protocol itself having long (or probabilistic) finality periods.

There are a number of protocols around today that can perform finality in <10 seconds, ours included, which allows us to have medium frequency trading.

This is probably enough for most users, hardcore day-traders need high frequency trading, which requires sub-second finality 😞

For that you're going to want to use traditional exchanges, or at best protocols with a very low #decentralization quotient.

3) Order books! #uniswap type #dex solutions are great, but they come with a bunch of limitations & issues.

To be a serious competitor to #cex dominance, a #dex must offer many of the same features. An order book is simply a must along with a trade matching engine.

A huge advantage an order book based #dex with a matching engine has is that a number of issues disappear such as MEV, front running and sandwich trades 🔥

4) Shorts 🧐 Order books allow short trades to be performed YAY! 🎉

However, there are a bunch of collateral/liquidity risks to consider if the short doesn't fill.

Solutions for those risks fall outside of the purely technical domain and are probably a headache 🤬

5) Stop losses 🤬 This one is a real nightmare & over the years I havent been able to find a viable solution for it. On a #dex your stop loss will be public.

The market knows where your stop loss is
and can potentially take advantage of that information. 😒

Ya'll can debate if that is a good thing or not as the market is now totally transparent.

I haven't revisited this issue for a few years though, advances in #zkps probably offers a solution. 🧐

6) Fiat off-ramp 🤬 These are still going to need a #centralized entity.

I did have some ideas for #social #decentralized #offramps but they were rife with KYC/AML issues.

Maybe there is a solution in the long term, but for now we're stuck with #centralized solutions. 😭

7) Conclusion is we're not *quite* there yet 😞

The main issues revolve around shorts, stop losses or any other trade classification where "the future" is involved.

Maybe it's fine to not support those features & day-traders remain on the #cex with everyone else inc retail using a #dex

Perhaps it's time to revisit these technical challenges that remain, tech capability here at #radix and elsewhere is moving forward quickly.

Either way, I don't think we will see #dex implementation that begin to challenge #cex implementations for a little while 🤬

Although, perhaps you can sleep a bit better at night knowing their days are probably numbered 👍